Machine Learning Principles Lecture by Hyerim Bae

These flashcards cover key concepts in machine learning principles discussed during the lecture.

Deductive Reasoning

Moving from general principles to specific instances.

Inductive Reasoning

Moving from specific instances to general principles.

Supervised Learning

Learning from training instances of known classification.

Unsupervised Learning

Learning from unclassified training data, involving conceptual clustering.

Reinforcement Learning

Learning optimal behavior in an environment to obtain the maximum reward.

Target Function

A function relating inputs to outputs, denoted as f: X → Y.

Hypothesis Set

A collection of hypotheses that can approximate the target function.

Gini Index

A measure of impurity of a dataset; values range from 0 (pure) to 1 (impure).

Entropy

A measure of the unpredictability or randomness; ranges from 0 to log2(m).

ID3 Algorithm

An algorithm to generate a decision tree for classification based on feature selection.

Perceptron Learning Algorithm (PLA)

An algorithm for supervised learning of binary classifiers.

Hoeffding's Inequality

An inequality that provides bounds on the probability that the sample mean deviates from the expected value.

VC Dimension

A measure of the capacity of a statistical model to classify data, representing the model's complexity.

In-sample error (Ein)

The error of a hypothesized model when fitted to the training data.

Out-of-sample error (Eout)

The error of a model when predicting for unseen data.